The Fed's hike is defensible. The years before it are not
Raising rates into a war premium, a tariff programme and an investment boom is the right call. The scandal is that it took all three, on top of a long overshoot, to produce it.
The Federal Reserve is America's central bank. It sets the interest rate that shapes what banks, businesses and households pay to borrow, and it aims to keep prices rising by about two per cent a year. Prices have been rising faster than that for several years. This week it raised the rate to slow them down.
Raising rates now is right, and upsetting the President is not a reason to avoid it. The real failure is the years of above-target inflation that came first, which will make this correction bigger than it needed to be.
The embarrassing number here is not the interest rate
A central bank that raises rates and irritates the head of government is, at minimum, doing something. The awkward fact in this week’s news is not the increase. It is the clause sitting behind it: the American economy has been dealing with above-target inflation for years. Not a quarter. Not a bad month for petrol. Years.
The Federal Reserve is the United States’ central bank. It sets the short-term interest rate that anchors what everyone else pays to borrow, and it has one numerical promise attached to its name: two per cent inflation over time. “Over time” is doing an enormous amount of work in that sentence. It is meant to buy the flexibility to ignore a one-off price jump. It was never meant to be a rolling licence.
Three causes, one remit
The reporting names three sources of pressure on prices: the aftermath of the war with Iran, the tariff programme, and the boom in artificial-intelligence investment. They are not the same kind of thing, and it is worth being clear about which ones a rate can touch.
An energy premium from a war is a relative price. A tariff is a tax on imports, and taxes raise the level of prices. Neither can be undone by monetary policy; the Fed cannot un-declare a conflict or rewrite a tariff schedule. The third is different. A construction and equipment boom — chips, buildings, power connections — is investment demand, and investment demand is exactly what an interest rate prices. That part sits squarely inside the remit, and it is the part that makes “these are all supply shocks, stand aside” a weaker line than it first appears.
What the Fed can always do, whatever the source, is decide how much of a shock is permitted to settle into wage bargains and price lists. That is the whole mechanism. It is a choice about expectations, not about oil barrels.
Why looking through has an expiry date
The case for ignoring a supply shock rests entirely on one condition: everyone continues to believe the target will be met later. That belief is what makes a manufacturer hesitate before passing a cost increase straight through, and what makes a union hesitate before building it into a three-year deal. Hesitation is the disinflation. It is free, and it comes from credibility rather than from unemployment.
Remove the belief and the arithmetic inverts. If the working assumption in a pricing meeting is that inflation will run above target, raising prices becomes the safe default rather than the risky one. The central bank then has to generate the hesitation the hard way, by pushing the real interest rate — the policy rate less expected inflation — far enough above neutral to slow hiring and spending. Every year of overshoot raises the price of the eventual correction. Delay does not avoid the tightening; it enlarges it.
The strongest case for sitting still
The best argument against this hike is genuinely good. Tariffs shift the price level once and then stop contributing to the rate of change. War premia in energy tend to fade as supply reroutes. An investment boom in computing capacity is disinflationary on the other side, once the capacity exists. On that reading, tightening now punishes a demand base that never misbehaved, and it bites just as the shocks are dropping out of the annual comparison — the classic sequence in which the medicine, not the disease, causes the downturn.
That case requires the shocks to be one-off and to arrive one at a time. These did neither. A tariff programme is standing policy that can be widened at will, which makes it a repeated level shift rather than a single one. A war is a condition, not an event. A capital-spending boom does not convert into productive capacity on the central bank’s timetable. Three overlapping pressures inside one window stop being a blip to be averaged away and become what the record already shows: a sustained miss. And note that the one-off defence was available in each of the preceding years too. An argument that is never wrong at the time and always wrong in hindsight is not a framework. It is a habit.
Irritation is not a cost
The expectation that this will rile the President is presented as a consequence of the decision. Read it instead as a test result. The point of taking rate-setting out of the hands of elected politicians is that the answer is sometimes unwelcome to whoever currently holds office. A decision that can be justified in monetary terms does not become wrong because it is politically inconvenient, and a central bank does not get to run loose policy to purchase quiet.
But independence has to be paid for with accountability, and this is where the American arrangement looks thin. On the same day, the Bank of England held its own rate and published an exchange of letters with the Chancellor of the Exchequer about consumer price inflation. Whatever those letters contain, the mechanism is the point: a miss produces a dated public document in which the Governor has to account for it in writing and the government has to respond in writing. The Fed’s misses produce testimony, minutes and press conferences. All useful. None of them a written answer to the only question that matters after a multi-year overshoot: why did this run so long, and what specifically changed your mind this month?
The Federal Open Market Committee should publish, alongside this decision, a short and explicit account of why it tightened now rather than a year ago — naming which measures of inflation expectations moved and by how much. Congress should go further and legislate a standing requirement modelled on the British practice: whenever inflation sits outside a stated band for a set number of consecutive months, the Chair writes a public letter explaining the miss and the path back, and the Treasury Secretary replies.
That costs nothing, constrains no decision, and gives the Fed something valuable in return. A central bank that has already documented its own lateness is much harder to bully over what it does next.
What would change this view
If this September increase turns out to be the last one, and twelve-month United States consumer price inflation is at or below two per cent by the Federal Open Market Committee's September 2027 meeting with no further rise in the policy rate in between, my case collapses. That path would show the overshoot was a sequence of genuine one-off shocks working through the numbers, that expectations never came unmoored, and that patience was the cheaper policy.
If that happens, the honest conclusion is that this hike was unnecessary insurance and that the critics of tightening into supply shocks were right on the substance. I would then be arguing about the wrong thing: not the delay, but the cost of an intervention that bought nothing the arithmetic would not have delivered anyway.
Worked from
- US Fed hikes interest rates to curb inflation in move likely to rile Trump
ABC News · 2026-09-16